Batch PreProcess problem

[mar504]

Using Batch PreProcessing Script v1.43 and it seems to be throwing errors regarding the master flat.

I have a bunch of lights and master dark/bias/flat frames. The first column of checkboxes on the Global Options side of the window are unchecked, the entire right column is checked (to use the master files). For some reason the lights are getting divided up into two subgroups when they are all the same type. Each says "Binning 1" and under that one group says "Position 1" and the other "Ha", here is an example of what it looks like with the first filename in each group.

Binning 1
  Position 1
    m57_1800sec_1x1_HA_2016-06-15_212722_-20.0C.fit
    m57_1800sec_1x1_HA_2016-06-16_210823_-20.0C.fit

Binning 1
  Ha
    m57_1800sec_1x1_HA_2016-06-22_220045_-20.0C.fit
    m57_1800sec_1x1_HA_2016-06-22_223116_-20.0C.fit

When I run the Diagnostics or try and run the script, I get the following errors. It seems as though it's not utilizing the flat frames. I've tried removing one of these mysterious groups of lights and itstill throws an error.

** Warning: Invalid file name characters will be replaced with underscores in filter name: 'Position 1'

** Warning: No flat frames have been selected to calibrate light frames (filter="Position 1", binning=1, exposure=1800.00s, length=11)

** Warning: No flat frames have been selected to calibrate light frames (filter="Ha", binning=1, exposure=1800.00s, length=9)

Seems like a bug, unless I'm doing something wrong?

[pfile]

it would seem that the script is unhappy with FITS keywords that have spaces in them... juan would have to look at this.

rob

[mar504]

Is there any way to downgrade to see if there is a problem with a specific version of the script?

[pfile]

there are old versions of the script here: https://github.com/PixInsight/PJSR/commits/master/src/scripts/BatchPreprocessing

but i suspect that this problem has been there forever. most people name their filters with names like O, H, R, or OIII, Ha, Red, etc. so its likely the script was never tested against filter names with spaces in them... but i'm just guessing here.

rob

[Juan Conejero]

No bug here, just normal behavior.

For some reason the lights are getting divided up into two subgroups when they are all the same type. Each says "Binning 1" and under that one group says "Position 1" and the other "Ha"

The reason is simple: There are frames acquired with different filters. A subset has been acquired with "Position 1" and the rest with "Ha". Filter names are retrieved automatically from standard FITS header keywords. Maybe "Position 1" frames have also been acquired with an Ha filter, but the script has no way to know that because the corresponding keyword says "Position 1". Obviously both names correspond to different filters, from the script's perspective.

A quick way to "fix" this problem is by adding frames manually. Click the Add Custom button and define the parameters as necessary.

On the other hand, where do the master frames come from? Have you generated them with the BPP script? Masters generated by other applications will not work.

Warning: Invalid file name characters will be replaced with underscores in filter name: 'Position 1'

This warning can be safely ignored. It simply says that the white space character in "Position 1" will be replaced with an underscore to build output file names. For compatibility reasons and to minimize potential problems with other tools and applications, the BPP script only uses alphabetic characters 'a-z' and 'A-Z', decimal digits '0-9', plus and minus signs '+-', and underscores '_'. Admittedly, the 'Invalid' word in this warning message should be 'Some'.

it would seem that the script is unhappy with FITS keywords that have spaces in them...

Not at all. The white space is a legal character and can be used without problems (although it is discouraged for different practical reasons, especially on platforms where one can have the pleasure to use a good terminal, such as FreeBSD and Linux and, to a lesser extent, OS X).

there are old versions of the script here

Obsolete versions of scripts and tools are not supported. Use them at your own risk and please don't report anything if you have problems.

[pfile]

fair enough, i am phoning it in these days,

rob

[mar504]

Thanks Juan, adding them with the custom button cleared up the messages. The masters were created with PI (manually, not with the BPP script). I think what may have happened is I hadn't defined the filter positions in SGP yet, so the default of "Position 1" was used for some, and at a later date when I acquired more subs I had changed the filter to "Ha", same filter but PI has no way of knowing that.